Different slopes for different folks: Variations in walkability-walking associations across sociodemographic groups

Past research suggests that walkability may be a weaker predictor of walking behavior for marginalized groups. While some studies have discussed potential reasons for this difference, few have tested them empirically. I address this gap using the 2017 National Household Travel Survey (n = 148,480), asking: (1) Do associations between walkability and walking differ across sociodemographic groups? (2) If so, what are the potential explanations? Using logistic regression models with interaction terms, I assess how associations between an objective walkability index (exposure) and utilitarian walking (outcome) vary by race/ethnicity and income, then examine how these differences are moderated by financial constraints and environmental barriers. The results show (1) weaker associations between walkability and walking for Black, Hispanic, and lower-income individuals and (2) differences in potential explanations for these results by race (infrastructure/safety disparities) vs. income (low vehicle access). These findings highlight the importance of differentiated strategies to promote walking across sociodemographic contexts.